What do molecular biologists mean when they say `structure determines function'? Gregor P. Greslehner∗ University of Salzburg & ERC IDEM, ImmunoConcept, CNRS/University of Bordeaux October 2018 Abstract `Structure' and `function' are both ambiguous terms. Discriminating different meanings of these terms sheds light on research and explanatory practice in molecular biology, as well as clarifying central theoretical concepts in the life sciences like the sequence{structure{function relationship and its corresponding scientific \dogmas". The overall project is to answer three questions, primarily with respect to proteins: (1) What is structure? (2) What is function? (3) What is the relation between structure and function? The results of addressing these questions lead to an answer to the title question, what the statement `structure determines function' means. ∗Email: [email protected] 1 Keywords: philosophy of biology, molecular biology, protein structure, biological function, scientific practice 1 Introduction `Structure' and `function' are abundantly used terms in biological findings. Frequently, the conjunct phrase `structure and function' or the directional phrase `from structure to function' is to be found, indicating that there is a special relation connecting these two concepts. The strongest form of this relation is found in the frequent statement that `structure determines function'. One could easily list several hundreds of references containing such phrases. However, in order not to blow up the references section, I will refrain from doing so. Suffice it to say that biologists make highly prominent use of these concepts in describing their research|molecular biologists, in particular. In this paper, I attempt to clarify these concepts, address their relation, and discuss the role they play in molecular biology's explanatory practice. While these issues can be addressed for many different biological entities on different levels of organization, I restrict the discussion primarily to proteins. What do biologists refer to when they use this phrase? Is there a particular scientific program or strategy behind the slogan `structure determines function'? Despite the frequent use of this phrase and the concepts to which it refers, a rigorous analysis is missing. Thus, a philosophical clarification would be a valuable contribution to the conceptual foundations of biology. One such fundamental concept is the sequence{structure{function relationship. \The relationships between sequence, structure, biochemical function and biological role are extremely ill-defined and scant 2 high quality data are available to allow us to analyse them." (Sadowski and Jones, 2009, 360) In this paper, I attempt to close this gap by developing an explication of both concepts of structure and function as they are used in biological practice and discussing which relation holds between them. The third component in this \trinity of molecular biology"|sequence|is the least in need of explication. The standard textbook view holds that sequence determines structure, and structure determines function. I will focus on the second relation. Without reviewing the rich history of these concepts throughout biology at this point, it is worth noting that functionality and form or structure were thought to be intimately linked from early on. In the early days of biology at the macroscale, the structures had to be observed with the naked eye. Thus, the first examples about the form of bodies or their parts and their functions can be found in physiology and anatomy, for example Harvey's notion of the heart's function to pump blood. From the scale of physiology to the molecular scale, structure and function are closely related. What exactly links these two concepts? Is it a determination relation? And if so, which one is determining the other? With the invention of microscopes and later the emergence of molecular biology, the structures and functions under consideration shifted from macroscopic entities to individual molecules. In fact, molecular biology put the three-dimensional shape of molecules center stage for explaining biological phenomena. This is the focus of this paper. In particular, the discussion will be confined to the structure and function of proteins|with special emphasis on the question whether the former determines the latter. 3 2 The ambiguity of `structure' In a first approximation, `structure' and `function' could be interpreted as the most general or neutral way of describing what molecular biologists are doing in their research and what their findings are about. These include mainly the three-dimensional shapes of molecules (or larger cellular structures) and the activities (functions) theses molecules perform in living cells, biochemical pathways, chemical reactions, or just individual steps in such mechanisms. The ultimate aim is to explain biological phenomena with molecular mechanisms, whose entities can be described in physical and chemical terms. The structure of molecules can be described in terms of physics and chemistry|function, however, is a concept that does not appear in physics or chemistry. Let's start by taking a closer look at the notion of structure. `Structure' is an ambiguous term. Applied to proteins, there is the usual nomenclature of primary structure (i.e., a protein's amino acid sequence), secondary structure (i.e., common structural motifs like α-helices and β-sheets), tertiary structure (i.e., the three-dimensional shape of a single folded amino acid chain), and quaternary structure (i.e., the final assembly of a protein if it consists of more than one amino acid chain). Other structurally important components are post-translational modifications and prosthetic groups which are not part of its amino acid composition. All these notions of structure have in common that they are about the molecular composition and shape of a molecule. One meaning of `structure' denotes the sequence of a polymer, the other meaning is about the three-dimensional shape of a molecule. As will be discussed below, another important ambiguity of `structure' allows to denote the organization of an interaction network. That leaves us with three different meanings of `structure': 4 (1) the sequence of a polymer, (2) the three-dimensional shape of a molecule, and (3) the network organization of several biological entities. While meanings (2) and (3) are candidates for being functional entities, structure as sequence (1) rather relates the sequences of different polymers (DNA, RNA, and proteins) and also plays a central role in determining the three-dimensional shape of a molecule, structure (2). The primary structure of a protein is just the sequence of amino acids that are put together to form a polypeptide. This amino acid sequence is determined by the corresponding protein-coding gene, which is first transcribed into mRNA and then translated into protein by the ribosome. This scheme is known as the \central dogma of molecular biology": DNA ! RNA ! protein The arrows might be interpreted as determination relations. The textbook view of protein structure and function proceeds as follows: nucleotide sequence ! amino acid sequence ! protein structure ! protein function Strong evidence supporting the claim that the three-dimensional shape of a protein is determined by the sequence of amino acids alone was provided by the experiments of Christian Anfinsen, showing that ribonuclease could, after treatment with denaturing conditions, regain its form and function (Anfinsen et al., 1961). Later, Merrifield showed that an in vitro synthesized sequence of amino acids can carry out the enzymatic activity of ribonuclease, thus gaining its functional form without the aid of any other cellular component (Gutte and Merrifield, 1971). From this and similar experiments, Anfinsen 5 built general rules of protein folding as a global energy minimum which depends solely on the sequence of amino acids (Anfinsen, 1973). This view is known as “Anfinsen’s dogma". In 1958, John Kendrew's lab determined the first actual three-dimensional form of a protein, myoglobin (Kendrew et al., 1958). The predominant technique to determine protein structures is still X-ray crystallography (Mitchell and Gronenborn, 2017). Other techniques include nuclear magnetic resonance, cryogenic electron microscopy, and atomic force microscopy. X-ray structures in particular have been supporting the view that there is a unique rigid shape|the protein's native, functional state|which would be necessary and sufficient for a protein to carry out its biological function. To make a long story short, the relation between nucleotide sequence and amino acid sequence has been generally confirmed (although there are much more complicated mechanisms to it, e.g., splicing). However, the part concerning the protein shape and function proves to be much more problematic. That poses a challenge to what Michel Morange calls \the protein side of the central dogma" (Morange, 2006). To get from amino acid sequence to three-dimensional structure is known as the protein folding problem. As the term `problem' suggests, it poses a serious challenge and remains unsolved to this day. Even though knowledge-based techniques to predict protein structures from their sequence have become impressively sophisticated, successful, and reliable, there are good reasons to suspect that the protein problem might remain unsolved in principle|if the aim is to predict protein folding based on chemical and physical principles only. Every two years the